x-agents-cross-review
Warn
Audited by Socket on May 1, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The review purpose is coherent, and most referenced tools fit that purpose, but the skill is over-privileged: it spawns multiple background agents with bypassPermissions and full tool access while processing untrusted PR/spec/code content. That makes prompt-injection and unintended autonomous actions the primary risks, with moderate data exposure risk and only low-to-moderate supply-chain concern from the underspecified curl usage.
Confidence: 89%Severity: 76%
Audit Metadata